Purpose

Cryogenically cooled coils have been a long-standing promise for SNR enhancement since the early days of MRI¹. This is even more relevant nowadays, with the increased interest on imaging of low abundance nuclei with lower Larmor frequencies than protons². However, the increased coil-to-sample distance needed in cryogenic coils can easily neutralize the SNR enhancement obtained by cooling. In this study we describe an MRI transparent cryostat for small animal imaging (Ø 73 mm) which allows a coil-to-sample distance below 3 mm, with the coil temperature reaching 88K while the outer surface being kept at room temperature.

Methods

A schematic of the cryostat is shown in Fig. 1. A separate ceramic (Al₂O₃) cylinder (AdValue, AZ, USA) whose outer diameter matches the inner diameter of the vessel acts as a cold-finger for to-be-cooled RF coil(s). Besides the vacuum thermal insulation, additional Multi-Layer Insulation (MLI) was installed, to reduce radiation heat losses. A combination of traditional (aluminized Mylar) and non-metallic (Kapton) MLI was used, in order to avoid eddy currents close to the RF coil (Fig. 2). The cryostat (ILK, Dresden, Germany) is shown in Fig. 3a, while the cold-finger (with the RF coil attached) is shown in Fig. 3b. Temperature sensors were attached to the cold-finger, and to the outside the cryostat (at the animal position). The temperatures were monitored under two different situations: without heat externally applied to the inner bore, and with warm air (50 ℃) applied. The approximate amount of LN₂ supplied during the experiment was 6 L. An octagonal 40x30 mm² RF coil made of flat copper (thickness 1.2 mm) was glued to the cold finger using thermally conductive epoxy, and its Q-factor measured at room temperature and when cooled down inside the cryostat. A reference measurement with the coil immersed in LN₂ was also performed. The coil includes tuning, matching and active decoupling as described in³. The SNR performance was measured with a combination of a spectrum analyzer (Keysight, CA, USA) and a signal generator (Rohde&Schwarz, Munich, Germany). A small pickup loop was placed inside the cryostat bore, at a distance of 38 mm to the coil, and excited with -100 dBm. The received signal was measured for two different cases: room temperature (290 K) and cooled (88K), and the SNR calculated.

Results

The measured temperature inside and outside the cryostat is shown in Fig. 4, where the warm air was applied for 5 min from minute 70 to 75, and later for a longer period of 50 min (from minute 100 to 150). It can be appreciated that the application of warm air, increases only minimally the temperature of the cold finger (0.6 K). The coil Q-factor is shown in Fig. 5.a). The Q-factor is increased from 325 to 599 when the coil is cooled, which agrees well with other results reported for coils of similar size^4,5. The measured SNR is shown in Fig. 5.b.

Discussion

The measured temperature at the coil position was 88 K (-184 °C), and this temperature could be kept stable for several hours (not shown in Figure 4). The temperature at the sample position was slightly decreased after some minutes to a stable value of about 10℃. This could be corrected by applying warm air at 50 ℃, with a minimal effect on the temperature of the cold-finger (and therefore on the coil). The relatively small difference between the measured Q-factor of the coil inside the cryostat, compared to the coil immersed in LN₂ shows that the coil is indeed very close to the LN₂ temperature (77 K). A 3-fold SNR enhancement was measured, which agrees well with the theoretically predicted value when sample losses are negligible $$$(ΔSNR = \frac{\sqrt{Q_{88K}·290}}{\sqrt{Q_{290K}·88}} = 2.6)$$$. The extra factor from 2.6 to 3 can be attributed to the noise reduction in the matching network, which was observed to effectively decrease the noise levels when cooled down.

Conclusion

Cryogenically cooled RF receive coils can be the means to significantly improve SNR in ¹³C MRI, where sample noise is not very important. This is so, as long as the coil-to-sample distance is not increased excessively due to the thermal insulation needed to provide a warm environment at the sample position. In this study we describe a small animal cryogenic coil setup, where the coil can be kept at a distance below 3 mm from the sample, while keeping the sample position at room temperature. A 3-fold SNR enhancement is reported for a small 30x40 mm² octagonal coil for ¹³C at 3T (32 MHz), which agrees well with the theoretically predicted value.

Acknowledgements

No acknowledgement found.